OpenCV之特征检测器（Feature Detector），描述子提取器（Descriptor Extractor）和描述子匹配器（Descriptor Matcher）

1.特征检测子

    -Harris

cv::cornerHarris(image,strength,3,3,0.01);

    -Fast

cv::Ptr<cv::FastFeatureDetector> fast = cv::FastFeatureDetector::create();

//或

cv::FAST(InputArray image, std::vector<KeyPoint> &keypoints, int threshold)

//或

cv::FAST(InputArray image, std::vector<KeyPoint> &keypoints, int threshold, bool nonmaxSuppression, int type)

    -SIFT

cv::Ptr<cv::xfeatures2d::SIFT> sift = cv::xfeatures2d::SIFT::create();

//或

<pre name="code" class="cpp">cv::Ptr<cv::xfeatures2d::SiftFeatureDetector> sift = cv::xfeatures2d::SiftFeatureDetector::create();

    -SURF

cv::Ptr<cv::xfeatures2d::SURF> surf = cv::xfeatures2d::SURF::create();

//或

cv::Ptr<cv::xfeatures2d::SurfFeatureDetector> surf = cv::xfeatures2d::SurfFeatureDetector::create();

    -ORB

cv::Ptr<cv::ORB> orb = cv::ORB::create();

    -MSER

cv::Ptr<cv::MSER> mser = cv::MSER::create();

    -GFTT

cv::Ptr<cv::GFTTDetector> gftt = cv::GFTTDetector::create();

或者直接用goodFeaturesToTrack function；

    -AGAST

cv::AGAST(InputArray image, std::vector<KeyPoint> &keypoints, int threshold)
//或
cv::AGAST(InputArray image, std::vector<KeyPoint> &keypoints, int threshold, bool nonmaxSuppression, int type)
//或
cv::Ptr<cv::AgastFeatureDetector> agast = cv::AgastFeatureDetector::create();

    -BRISK

cv::Ptr<cv::BRISK> brisk = cv::BRISK::create();

    -SimpleBlob

cv::Ptr<cv::SimpleBlobDetector> blob = cv::SimpleBlobDetector::create();

    -KAZE

cv::Ptr<cv::KAZE> kaze = cv::KAZE::create();

    -AKAZE

cv::Ptr<cv::AKAZE> akaze = cv::AKAZE::create();

2.描述子提取器

    -SIFT

<span style="font-weight: normal;"><span style="font-size:12px;">cv::Ptr<cv::xfeatures2d::SIFT> sift = cv::xfeatures2d::SIFT::create();
//或
cv::Ptr<cv::xfeatures2d::SiftDescriptorExtractor> sift = cv::xfeatures2d::SiftDescriptorExtractor::create();</span></span>

    -SURF

<span style="font-weight: normal;"><span style="font-size:12px;">cv::Ptr<cv::xfeatures2d::SURF> surf = cv::xfeatures2d::SURF::create();
//或
cv::Ptr<cv::xfeatures2d::SurfDescriptorExtractor> surf = cv::xfeatures2d::SurfDescriptorExtractor::create();</span></span>

    -BRIEF

Opencv中没有单独将BRIEF用来提取描述子，因为它是被用于ORB中的；

    -BRISK

<span style="font-weight: normal;"><span style="font-size:12px;">cv::Ptr<cv::BRISK> brisk = cv::BRISK::create();</span></span>

    -ORB

<span style="font-size:12px;font-weight: normal;">cv::Ptr<cv::ORB> orb = cv::ORB::create();</span>

ORB是用fast特征； 

   -KAZE

<span style="font-size:12px;font-weight: normal;">cv::Ptr<cv::KAZE> kaze = cv::KAZE::create();</span>

注：KAZE描述子只能使用KAZE或AKAZE特征点；

    -AKAZE

<span style="font-size:12px;font-weight: normal;">cv::Ptr<cv::AKAZE> akaze = cv::AKAZE::create();</span>

同样，AKAZE描述子也只能使用KAZE或AKAZE特征点；

3.描述子匹配器

   -BruteForce-BFMatcher

cv::Ptr<cv::BFMatcher> bf = cv::BFMatcher::create("BruteForce");

其中，匹配类型还可以是“BruteForce-L1”，“BruteForce-L2”，“BruteForce-Hamming”；

   -FlannBased-FlannBasedMatcher

cv::Ptr<cv::FlannBasedMatcher> flann = cv::FlannBasedMatcher::create("FlannBased");